CNC controller

The interface is pretty simple. A PC running control software handles the logical side.
The controller will move the motors per the PC's commands. An old PC case is handy because the power supply can be modified to power the drivers and the motors.

The parallel port on the pc will signal step and direction - the driver will actually step the motor in the correct direction by alternating current and voltage to the motors coils.
For the controller, I've chosen to use the Microstep design from EAS.
Using the 8 microstep mode with my 1.8 degree steppers (200 full steps/rev)
I get 1600 steps per revolution. Combined with .2"/revolution ball screws
I'll actually achieve .000125" per microstep.
Manual Schematic layout Parts list
I ordered three microstep boards and a parallel breakout board from EAS.
I've got several of the driver chips - here's the parts list I ordered from digi-key. The list on the microstep page is a little out of date. I had to adjust some quantities and change some part numbers for availability as of 3/07.

The controller is built into a small pc case
Each microstep has it's own heat sink.
Each one gets power directly from the power supply
The spare is for adding a fourth axis later on.

wiring

The yellow wires provide 12 volts to the motors
The black is ground - which is the same bus on
the microsteps, so only one wire is needed
The red wire providies 5v directly to the pic
circuitry. This bypasses the 7805s onboard.
The speakerwire was the only thing thick enough
to carry full current to the stepper motors
Control signals come into the breakout board.
A DB-25 straight through cable connects to the pc.
I forgot to order the resistor network, so I made one
The power supply is a smaller form factor ATX psu.
The white part is a 271-120 20 watt 8 ohm resistor from radioshack.
It's connected to the 5 volt bus to force the psu to produce
enough current on the 12V bus.
The power switch is a spare I had laying around.
It's mounted to the front cover of the pc case.
It's wired to the green psu lead (power enable)
and one of the ground leads. When the green wire is grounded
by the switch, the power supply is turned on.

Stepper Motors

I picked up a pair of 190oz/240oz (unipolar/bipolar) Vexta PK268-E2.0A
stepper motors for the X and Y axis off ebay. The specs are in the datasheet, but I'll run down the key bits here.
I lucked out with these motors. I bought them with the idea that they'd need 2.0 amps of current. But to run them in Bipolar parallel mode, they'll actually need 2.8 Amps per phase. The microstep controller is rated for 3.0 amps, making these an ideal setup.
Mode torqueCurrent
Bipolar parallel240oz-in2.8A/phase
Bipolar series240oz-in1.4A/phase
Unipolar190oz-in2A/phase
The motors I picked up happen to be 8 wire - making them very flexible. The graph above explains why I'd prefer the bipolar-series arrangement - the motors will have maximum torque and maintain it into higher RPMs. (Allowing the machine to operate faster)

I'll get a Z motor once I decide how to modify the axis.
To couple the motors to the ball screws,
I wanted some serious couplers that would avoid binding

I ordered my lovejoy spider couplings
from Enco. OD=1.08", OAL=1.72"
I bought 1/4" hole hubs, and Buna-n spiders
Part numbers:
990-4044 (hubs) x4
990-4042 (spiders) x 2

Useful reading/links

A very good paper on microstepping vs. full step torque. There are some trade offs, but it's less of an issue than you'd think.